Purpose <p>To determine which parameter best identifies those at risk of clinically significant postoperative residual astigmatism among eyes classified as low-astigmatism by SimK.</p> Methods <p>Patients with preoperative simulated keratometric (SimK) astigmatism &lt; 0.75 D were included. All participants underwent phacoemulsification and monofocal intraocular lens (IOL) implantation. Preoperative SimK astigmatism, total keratometric (TK) astigmatism, predicted refractive astigmatism (PRA) calculated by Barrett Toric with predicted posterior corneal astigmatism (PCA) and measured PCA were evaluated with a 0.75D indication for astigmatism correction. Receiver Operating Characteristic (ROC) curves were further constructed to determine whether a suitable indication was better than 0.75D. Sensitivity analysis was also conducted with 0.50 D as a threshold value.</p> Results <p>In total, 723 patients (723 eyes) were included with a mean age of 63.34 ± 10.26 years. Of these, 275 eyes (38.04%) had postoperative refractive astigmatism ≥ 0.75 D and 504 eyes (69.71%) had residual astigmatism ≥ 0.50 D when SimK was applied in preoperative evaluation. Using TK astigmatism and PRA yielded significantly lower mean centroid and mean absolute prediction errors than SimK astigmatism (<i>P</i> &lt; 0.05). ROC curves indicated that by lowering the threshold value (range: 0.45-0.64D), the sensitivity of PRA and TK astigmatism raised significantly (TK: from 39.27% to 62.73%, PRA: from 50.55% to 65.45%).</p> Conclusions <p>Among eyes preoperatively classified as low-astigmatism by SimK, PRA and TK showed better risk-stratification performance than SimK. Astigmatism correction should be considered when PRA or TK values exceed 0.6 D.</p> Translational relevance <p>The threshold of TK astigmatism and PRA can serve as a quick reference for risk identification of residual astigmatism.</p>